This invention is directed to biomaterials for spatially and temporally controlled delivery of bioactive agents such as drugs, growth factors, cytokines or cells. In particular, this invention teaches versatile methods for chemical crosslinking of high molecular weight hyaluronic acid under physiological conditions in situ, to form polymerizable biodegradable materials. The methods are based on the introduction of functional groups into hyaluronic acid (HA) via formation of an active ester at the carboxylate of the glucuronic acid moiety as an intermediate and subsequent substitution with a side chain containing a nucleophilic group on one end and a (protected) functional group on the other end. The introduced functional groups allow for crosslinking of the HA derivatives. Crosslinked hyaluronic acid hydrogels of this invention are useful in various surgical applications and as a temporary scaffold for tissue regeneration, e.g., in cartilage repair.
The failure of regenerating persistent hyaline cartilage by surgical procedures has prompted investigators to attempt repair using biological strategies. The biological repair of articular cartilage is, with a few exceptions, still at an experimental stage. Biological cartilage repair has been approached in two basic ways. First, autologous chondrocytes have been transplanted into a lesion to induce repair (Grande et al., J. Orthop. Res. 7, 208-214 (1989); Brittberg et al., New Engl. J. Med. 331, 889-895 (1994); Shortkroffet al., Biomaterials 17, 147-154 (1996)). Chondrocytes may be obtained from a low-loaded area of a joint and proliferated in culture (see Grande; Brittberg; Shortkroff, supra), or mesenchymal stem cells may be harvested, e.g., from the iliac crest marrow, and induced to differentiate along the chondrocyte lineage using growth factors (Harada et al., Bone 9, 177-183 (1988); Wakitani et al., J. Bone Joint Surg. 76-A, 579-592 (1994)). The chondrocyte transplantation procedures currently attempted clinically, although promising, are hampered because technically they are very challenging, the cell preparation is very expensive, and the potential patient pool is limited by age, defect location, history of disease, etc. Cells have also been transplanted into cartilage defects in the form of perichondral grafts, e.g., obtained from costal cartilage, but with limited success due to the limit in donor material and the complication of endochondral ossification of the graft site observed in longterm follow-up (Amiel et al., Connect. Tissue Res. 18, 27-39 (1988); O""Driscoll et al., J. Bone Joint Surg. 70-A, 595-606 (1988); Homminga et al., Acta Orthop. Scand. 326-329 (1989); Homminga et al., J. Bone Joint Surg. 72-B, 1003-1007 (1990)). A second approach is aimed at the recruitment of mesenchymal stem cells from the surrounding connective tissue, e.g., synovium, using chemotactic and/or mitogenic factors (Hunziker and Rosenberg, J. Bone Joint Surg. 78-A, 721-733 (1996); see also U.S. Pat. No. 5,368,858). The availability of growth factors and cytokines in recombinant form and the lack of complicated cell transplantation make this procedure a very attractive alternative. The shortcoming of both procedures is the difficulty to stably anchor the repair-inducing factors, whether tissue grafts, cells, or growth factors, within the defect site. Also, outlining of the space that is to be repaired, e.g., by filling it with a matrix material, appears to be crucial to recreate a level cartilage surface (Hunziker and Rosenberg, supra). Thus far, the availability of candidate matrix materials has been the limiting factor, and anchoring of materials seeded with chondrocytes and/or chondrogenic factors difficult, explaining the unsatisfactory results obtained with currently available materials such as polylactic acid and polyglycolic acid scaffolds (Freed et al., J. Biomed. Mat. Res. 28, 891-899 (1994); Chu et al., J. Biomed. Mat. Res. 29, 1147-1154 (1995)); calcium phosphate minerals (Nakahara et al., Clin. Orthop. 276, 291-298 (1992)), fibrin sealants (Itay et al., Clin. Orthop. 220, 284-303 (1987)), and collagen gels (Wakitani et al., J Bone Joint Surg. 71-B, 74-80 (1989)). We have developed novel biodegradable materials based on hyaluronic acid which are optimized for the biological requirements posed on a repair material in a synovial joint and which allow in situ polymerization.
Hyaluronic acid (HA) is unique among glycosaminoglycans in that it is not covalently bound to a polypeptide. HA is also unique in having a relatively simple structure of repeating nonsulfated disaccharide units composed of D-glucuronic acid (GlcUA) and N-acetyl-D-glucosamine (GlcNAc) (Scott et al., The Chemistry. Biology and Medical Applications of Hyaluronan and Its Derivatives, T. C. Laurent (ed.), Portland Press, London, (hereinafter xe2x80x9cHyaluronan and Its Derivativesxe2x80x9d), pp. 7-15 (1998)). Its molecular mass is typically several million Daltons. HA is also referred to as hyaluronan or hyaluronate, and exists in several salt forms (see formula I). 
HA is an abundant component of cartilage and plays a key structural role in the organization of the cartilage extracellular matrix as an organizing structure for the assembly of aggrecan, the large cartilage proteoglycan (Laurent and Fraser, FASEB J. 6, 2397-2404 (1992); Mxc3x6rgelin et al., Biophys. Chem. 50, 113-128 (1994)). The noncovalent interactions of aggrecan and link protein with HA lead to the assembly of a large number of aggrecan molecules along the HA-chain and mediate retention of aggrecan in the tissue. The highly negatively charged aggrecan/HA assemblies are largely responsible for the viscoelastic properties of cartilage by immobilizing water molecules. A number of cell surface receptors for HA have been described and shown to play a critical role in the assembly of the pericellular matrix of chondrocytes and other cells, e.g., isoforms of CD44 and vertebrate homologues of Cdc37 (Knudson and Knudson, FASEB J. 7, 1233-1241 (1993); Grammatikakis et al., J. Biol. Chem. 270, 16198-16205 (1995)), or to be involved in receptor-mediated endocytosis and degradation of HA to control HA levels in tissues and body fluids (Laurent and Fraser, supra; Fraser et al., Hyaluronan and Its Derivatives, pp. 85-92 (1998)). Blocking of the interaction of these receptors with HA in prechondrogenic micromass cultures from embryonic limb bud mesoderm inhibits chondrogenesis, indicating that the establishment and maintainance of a differentiated chondrocyte phenotype is at least in part dependent on HA and HA-receptor interactions (Maleski and Knudson, Exp. Cell. Res. 225, 55-66 (1996)).
HA and its salts are currently being used in therapy for arthropathies by intraarticular injection (Strachnan et al., Ann. Rheum. Dis. 49, 949-952 (1990); Adams, Hyaluronan and Its Derivatives, pp. 243-253 (1998)), in opthalmic surgery for intraocular lens implantation (Denlinger, Hyaluronan and Its Derivatives, pp. 235-242 (1998), to promote wound healing in various tissues (King et al., Surgery 109, 76-84 (1991)), or more recently, in derivatized and/or crosslinked form to manufacture thin films which are used for tissue separation (for review see Laurent and Fraser, supra; Weiss, Hyaluronan and Its Derivatives, pp. 255-266 (1998); Larsen, Hyaluronan and Its Derivatives, pp. 267-281 (1998); Band, Hyaluronan and Its Derivatives, pp. 33-42 (1998)). Extensive efforts have been made by various laboratories to produce derivatives of HA with unique properties for specific biomedical applications. Most of the developments have been focusing on the production of materials such as films or sponges for implantation and the substitution of HA with therapeutic agents for delayed release and/or prolonged effect (for review see Band, supra; Prestwich et al., Hyaluronan and Its Derivatives, pp. 43-65 (1998); Gustafson, Hyaluronan and Its Derivatives, pp. 291-304 (1998)). Strategies have included esterification of HA (U.S. Pat. Nos. 4,957,744 and 5,336,767), acrylation of HA (U.S. Pat. No. 5,410,016), and cross-linking of HA using divinyl sulfone (U.S. Pat. No. 4,582,865) or glycidyl ether (U.S. Pat. No. 4,713,448). However, the modified HA molecules show altered physical characteristics such as decreased solubility in water and/or the chemical reaction strategies used are not designed for crosslinking of HA under physiological conditions (in an aqueous environment, at pH 6.5-8.0).
It is well known that polyaldehydes can be generated by oxidizing sugars using periodate (Wong, CRC Press, Inc., Boca Rayton, Fla., pp. 27 (1993); European Patent No. 9615888). Periodate treatment oxidizes the proximal hydroxyl groups (at C2 and C3 carbons of glucuronic acid moiety) to aldehydes thereby opening the sugar ring to form a linear chain (Scheme 1). While periodate oxidation allows for the formation of a large number of functional groups, the disadvantage is the loss of the native backbone structure. Consequently, the generated derivative may not be recognized as HA by cells. In fact, hydrogels formed by using periodate oxidized HA as a crosslinker, e.g., in combination with the HA-amines described herein, showed very limited tissue transformation and poor cellular infiltration in the rat ectopic bone formation model (FIG. 6). This is in sharp contrast to the HA-aldehyde derivatives described herein.
The introduction of free amino groups on HA, which could be used for further convenient coupling reactions under mild physiological conditions, has been a subject of great interest. Previous methods have produced a free amino group on high molecular weight HA by alkaline N-deacetylation of its glucosamine moiety (Curvall et al., Carbohydr. Res. 41, 235-239 (1975); Dahl et al., Anal. Biochem. 175, 397-407 (1988)). However, concomitant degradations of HA via beta-elimination in the glucuronic acid moiety was observed under the harsh reaction conditions needed. This is of particular concern because low molecular weight HA fragments, in contrast to high molecular weight HA, have been shown to be capable of provoking inflammatory responses (Noble et al., Hyaluronan and Its Derivatives, pp. 219-225 (1998)). An early report claimed that carbodiimide-catalyzed reaction of HA with glycine methyl ester, a monofunctional amine, led to the formation of an amide linkage (Danishefsky and Siskovic, Carbohydr. Res. 16, 199-201 (1971)). This however, has been proven by a number of studies not to be the case (Kuo et al., Bioconjugate Chem. 2, 232-241 (1991); Ogamo et al., Carbohydr. Res. 105, 69-85 (1982)). Under mildly acidic conditions the unstable intermediate O-acylisourea is readily formed, which in the absence of nucleophiles, rearranges by a cyclic electronic displacement to a stable N-acylurea (Kurzer and Douraghi-Zedeh, Chem. Rev. 67, 107-152 (1967)). This Oxe2x86x92N migration of the O-acylisourea also occurs when the nucleophile is a primary amine (Kuo et al., supra) and any amide formation that does occur is insignificant as reported by Ogamo et al., supra. Experiments where high molecular weight HA (Mrxcx9c2xc3x97106 Da) was reacted with an excess of the fluorescent label 5-aminofluorescine in the presence of the carbodiimide EDC achieved only 0.86% of theoretical labelling. The introduction of a terminal hydrazido group on HA with a variable spacer has recently been achieved and has led to the ability to conduct further coupling and crosslinking reactions (Pouyani and Prestwich, Bioconjugate Chem. 5, 339-347 (1994), U.S. Pat. Nos. 5,616,568, 5,652,347, and 5,502,081; Vercruysse et al., Bioconjugate Chem. 8, 686-694 (1997)).
It is an objective of this invention to provide a method for more versatile modification of HA with various functional groups that allow for crosslinking of the HA derivatives under physiological conditions. It is another objective that the method of functionalization does not compromise the molecular weight or chemical identity (except of the target carboxyl group for coupling) of HA. It is a further objective that the method of functionalization provides HA molecules that are well tolerated in vivo and are biodegradable.
It is also an objective of this invention to identify HA derivatives and methodology for in situ polymerization thereof to provide a biodegradable scaffold for tissue regeneration. It is another objective that the HA materials can be polymerized in the presence of cells to serve as a vehicle for cell transplantation. It is a further objective to provide methodology for functionalization and cross-linking of HA that allows for variations in the biomechanical properties of the formed gels as well as in the sensitivity to cellular infiltration and degradation.
Biomaterials for spatially and temporally controlled delivery of bioactive agents such as drugs, growth factors, cytokines or cells, are a key factor for tissue repair. In particular, in situ polymerizable biodegradable materials are needed for cartilage resurfacing that are designed to withstand the mechanical forces in a joint. We have developed a versatile method for chemical crosslinking of high molecular weight hyaluronic acid under physiological conditions. The method is based on the introduction of functional groups into hyaluronic acid by formation of an active ester at the carboxylate of the glucuronic acid moiety and subsequent substitution with a side chain containing a nucleophilic group on one end and a (protected) functional group on the other end. We have formed hyaluronic acid with amino or aldehyde functionality, and formed hydrogels with modified hyaluronic acid and bifunctional crosslinkers or mixtures of hyaluronic acid carrying different functionalities using active ester- or aldehyde-mediated reactions. Physical and chemical properties of the hydrogels of this invention were evaluated using biomechanical testing, and by assaying sensitivity towards degradation by glycosidases such as testicular hyaluronidase. Biocompatibility was evaluated using cell culture assays and subcutaneous implantation of the hyaluronic acid materials in rats. This in vivo assay is also the established model for induction of ectopic bone formation by members of the transforming growth factor xcex2 family (TGF-xcex2), and several crosslinked hyaluronic acid materials of this invention gave excellent ectopic bone formation in vivo when loaded with appropriate growth factor(s).
As set forth below in the detailed description of the invention, the compositions of the invention have many therapeutic uses. For example, compositions of the invention may be used to stem hemorrhage in general surgery, reconstruct nerves and vessels in reconstructive, neuro- and plastic surgery, and to anchor skin, vascular, or cartilage transplants or grafts in orthopedic, vascular, and plastic surgery. Compositions of the invention are also useful as vehicles for the delivery of cells or bioactive molecules such as growth factors to stimulate focal repair. Local delivery of growth factors facilitates wound healing and tissue regeneration in many situations, not only in promoting bone formation and stimulating cartilage repair in orthopedic procedures, but also, e.g., in treating pathological wound conditions such as chronic ulcers. These compositions may also serve as a scaffold to generate artificial tissues through proliferation of autologous cells in culture. On the other hand, the anti-adhesive property of some compositions with respect to cells render such compositions particularly suitable to generate tissue separations and to prevent adhesions following surgery. The viscoelastic properties of HA make it particularly well suited for this purpose, and it is used clinically to achieve temporal pain relief by repeated intraarticular injections in arthropathies as a xe2x80x9cjoint lubricantxe2x80x9d, as a protective agent for eye irritations and in ophthalmic surgery, as a barrier to cells in facial and other reconstructions in plastic surgery and dentistry, in reconstructive surgery of tendons, in surgical procedures in the urogenital system, and in thoracic surgery. The injectable nature of the compositions of the invention also renders them suitable for tissue augmentation in plastic surgery, where the HA matrix serves primarily as an inert biocompatible filler material (Balasz and Laurent, Hyaluronan and Its Derivatives, pp. 325-326 (1998)), e.g., for filling dermal creases or lip reconstruction.
HA hydrogels match several of the desired properties for a biodegradable material biocompatible with cells. The relatively simple repetitive structure of HA allows for specific modification and introduction of a large number of functional groups, for crosslinking to generate hydrogels with excellent physical properties. HA hydrogels have also successfully been used as a delivery vehicle in chondrocyte transplantation studies (Robinson et al., Calcif Tissue Int. 46, 246-253 (1990)) and HA has proven its biocompatibility in various forms in clinical practice (for review see Laurent and Fraser, supra; Balazs and Laurent, supra).
The reaction mechanisms we have explored for in situ polymerization of HA derivatives are compatible with an aqueous environment and are non-toxic to cells. The aldehyde-mediated crosslinking strategies follow reactions occurring physiologically in crosslinking of fibrillar collagens and elastin. NHS-esters provide an alternative for rapid formation of stable bonds under physiological conditions, primarily by reaction with primary amines. The technology of NHS-ester-mediated protein crosslinking has been developed for materials with applications in plastic surgery that require in situ polymerization (U.S. Pat. No. 5,413,791)).